Token Economics Waterline

Token Economics Waterline is the practice of calculating, before any client work is scoped, the exact per-million-token cost at which a given model stops protecting the retainer margin. Foundation model platforms price per token and meter by tier, so a workflow that looks cheap in a demo can cross the waterline once real client volume, retries and long context windows are applied. The framework asks three questions per build: what is the blended input/output cost at projected monthly volume, how much headroom exists before the retainer margin drops below target, and which cheaper tier or shorter context keeps the work above the line. Routing decisions inside mixture-of-experts models change inference cost and latency for identical prompts, so the same task can sit on either side of the waterline depending on which model tier an agency routes it to. Run the arithmetic on a live client workflow, not a benchmark.